Junior Applied AI Engineer (LLM Systems)

About the role

Coforge is hiring a Junior Applied AI Engineer focused on LLM systems, working remotely within LATAM. The role covers building and evaluating AI agents, guardrails and moderation systems, and improving prompts and workflows, in collaboration with Engineering, Data Science and Compliance teams.

Key responsibilities

• Design and run evaluation frameworks for AI agents, including test datasets, LLM-as-a-Judge methods, human reviews and regression testing.
• Build, configure and optimise AI guardrails and moderation systems for regulated domains such as financial services and credit products.
• Develop, test and improve prompts based on evaluation results and production trace analysis.
• Create and maintain MCP (Model Context Protocol) servers and tool integrations so AI agents can access internal data and services.
• Support fine-tuning and evaluation of language models with domain-specific datasets.
• Prototype and validate AI workflows using System One models such as Jev and Laya for classification, routing and verification.
• Monitor AI system quality, latency, performance and costs through Langfuse and drive improvements.
• Work with Engineering, Data Science and Compliance to keep AI solutions reliable, secure and compliant.

Required profile

• 1-3+ years of experience in AI/ML engineering, applied AI, data science or software engineering.
• Experience with large language models and AI-powered applications.
• Knowledge of LLM evaluation techniques, including benchmark datasets, scoring frameworks and LLM-as-a-Judge approaches.
• Understanding of AI agent behaviour in multi-turn conversations and tool-calling workflows.
• Advanced English, written and spoken.

Required skills

• Prompt engineering, structured outputs and prompt optimisation.
• AI safety concepts: moderation, prompt injection prevention, jailbreak mitigation and guardrails.
• AWS cloud services, preferably Amazon Bedrock.
• Langfuse or similar observability and evaluation platforms.
• Python and REST APIs with FastAPI or similar.
• Unit testing, code reviews and Git workflows.
• Preferred: MCP server development, LoRA or full fine-tuning, small language models, System One models, and regulated industry experience.